Post on 08-Aug-2020
Title page Dialogue Design - With Mutual Learning as Guiding Principle
Submitted version1
International Journal of Human-Computer Interaction, "Readings from NordiCHI2000" (special issue), Lawrence Erlbaum Associates, Inc (2003). , Vol. 15, No. 1, Pages 21-40 Janni Nielsen, Copenhagen Business School, Department of Informatics Lone Dirckinck-Holmfeld, Aalborg University, Department of Communication Oluf Danielsen, Roskilde University, Department of Communication Running head title: Dialogue Design Biographical information:
Janni Nielsen, PhD in Informatics and Psychology is an Associate Professor in HCI
at Copenhagen Business School. Her research focus is visual interaction, design of
interfaces and user involvement methods.
Lone Dirckinck-Holmfeld, PhD in communication is Research Professor conducting
research and development work within the area computer supported collaborative
learning (CSCL)
Oluf Danielsen, PhD in physics is an Associate Professor in Communication at
University of Roskilde with particular research activities within ICT and
communication in organizations; user involvement methods.
1 Published version: http://www.leaonline.com/doi/pdf/10.1207/S15327590IJHC1501_03?cookieSet=1
Dialogue Design
- With Mutual Learning as Guiding Principle
Janni Nielsen
Copenhagen Business School, Department of Informatics
Lone Dirckinck-Holmfeld
Aalborg University, Department of Communication
Oluf Danielsen
Roskilde University, Department of Communication
ABSTRACT
The paper describes a large European R&D-project on Multimedia And Network
In Co-operative Research And Learning (MANICORAL) from point of view of the
participating HCI-researchers. The project developed the methodology of Dialogue
Design, drawing on two sources: Participatory Design and Dialogue Research. Action
Research is understood as the historical basis for the two strands, , where
Participatory Design has focused on research in working life, and Dialogue Research
has focused on living conditions. However, Dialogue Design as a methodology differs
from the above in a number of aspects In Dialogue Design the carrying principal is
mutual learning, focus is on working life of high resource groups, and these users are
themselves developing parts of the technologies. The techniques applied and the role
of the HCI-researcher as mediator creating fora for dialogues are introduced and
reflected upon, and Dialogue Design is discussed within the theoretical concepts of
communication and learning.
CONTENTS
1. INTRODUCTION
1.1 The Expert Users
1.2 The Human Computer Interaction Group (HCI-Group)
2. SETTING THE SCENE
3. FROM ACTION RESEARCH TO DIALOGUE RESEARCH AND
PARTICIPATORY DESIGN
3.1 Working Conditions
3.2 Living Conditions
4. DIALOGUE DESIGN
4.1 Dialogue - A Mutual Learning Process
5. PRACTICES
5.1 Understanding Practice - Baseline Data and Requirements Capture
5.2 Visionary Practice - Forays of Dialogues
5.3 Decision Techniques - From Requirements to Specification
6. MUTUAL INTERVIEWING AS A TOOL FOR DIALOGUE AND
LEARNING
7. DIALOGUE DESIGN AS MUTUAL LEARNING - SUMMING UP ON
THE PRINCIPLES
8. REFLECTIONS ON THEORETICAL FRAME
1. INTRODUCTION
MANICORAL (Multimedia And Network In Co-operative Research And
Learning) was an international R&D-project, involving six European nationalities. It
was organised as an interdisciplinary project with participants from four scientific
disciplines: the Natural Sciences, the Social Sciences, Technology and the
Humanities. The duration of the project was two years supported by the European
Union's 4th Frame program: Telematics for Research (1996-1998).
The aim of the project was to develop a Distributed Collaborative Visualisation
system (DCV) for a dispersed group of European scientists. The users were
researchers from different disciplines within the area of Natural Sciences. The system
developers were experts in visualisation and collaboration systems and the HCI-group
had its background within the Social Sciences (sociology) and the Humanities
(communication and psychology). With basis in the expert user group, this paper
presents and discusses the perspective of the HCI-participants.
1.1 The Expert Users
The users were a group of geophysicists investigating the use and exploitation
of Radar Altimetry Data, and their project on Altimetry for Research In Climate And
Resources (AFRICAR) was supported by the European Space Agency. The scientists
were located in Holland, Austria, Italy, Greece and Denmark and had collaborated for
many years through conventional technology and face-to-face meetings. Their
research aim was (and is) to utilise methods of measuring distances from satellites to
the ocean surface or to the ice surface. Change of the altitude of the ice cap may
indicate climate change. Studying the variations of ocean currents and eddies may
contribute to a basic understanding of climate change, and the observations are also
used for calculation of the gravity field.. This is important for mapping resources, e.g.
oil and for understanding/predicting e.g. earthquakes. (Nielsen, Lindgaard, Dirckinck-
Holmfeld, Vendelø, Danielsen & Georgsen, 1997).
Though all the domain experts were geophysicists, they belonged to five
different research communities with each their individual expertise: theoretical
mathematicians from Italy; specialists in ocean tides and circulation from Denmark;
satellite position and orbit experts from Austria and Holland, the latter also being
experts in computer generated scientific visualisations and imations (image and
motion). Greece represented the community of application experts and a Danish
University community supplied the gravity experts.
1.2 The Human Computer Interaction Group (HCI-Group)
The HCI-group had a number of tasks in the project. One of which was to
contribute to the development of a methodological and theoretical framework for
investigating collaboration between knowledge workers in a distributed context. The
group was also responsible for studying – on a long-term basis – the constitutive
influence of the technology on collaboration, and on the local communities. The EU
only supplied funds for two years of the initial four years of the project proposal. Due
to the lack of funding, the project ended and the long term studies had to be given up
although the core group of HCI-researchers (the authors of this article) have continued
a low key research co-operation.
As the basis for the project, the HCI-group had the role of mediators. They were
responsible for initiating interaction among participants, creating room for dialogue,
ensuring procedures for collaboration and acting as scribes – keeping track of the
development. The challenge for the HCI-researchers was, as indeed for all
MANICORAL-participants, how to get this very interdisciplinary, inter-cultural,
international and distributed community - constituted by many different groups - to
work together and to design a distributed visualisation tool. With such diversity not
only between scientific disciplines, nationalities and cultures but also within the
domain groups, it became essential to reach a common ground.
Figure 1: Illustration of the interdisciplinary and the international dimensions of
MANICORAL
"FIGURE 1 INSERT HERE"
A multi-cultural project requires very careful consideration. Our concern was
how to ensure co-operation and communication so that the participants would be able
to work out solutions together, build on each others' ideas, become visionary and
design together. But, more than anything, the important questions to be considered
were how to ensure that they would:
- come to understand each other and develop mutual respect?
- come to trust each other?
- engage in mutual learning processes?
- share knowledge and understandings?
Looking back, it was a very complex process, like most large European design
projects are. However, MANICORAL resulted in a first step in the direction of
answering the above questions. The project became rooted in the methodology of
Dialogue Design (DD) in which the carrying principle is mutual learning. In the
following, the methods applied will be introduced and reflected upon, initially by
placing them within the frames of Action Research.
2. SETTING THE SCENE
Action Research is traditionally traced back to the first half of the 20th century
and to the classic organisation studies in the 20ies of the Hawthorne factories under
Western Electric in Chicago. During the 2nd World War, the German-American
social psychologist Kurt Lewin (Lewin, 1975) conducted Action Research in the U.S.
on a series of projects on poor and minority groups. Historically, Action Research
became connected with underprivileged groups. The research findings would
(ideally) serve as basis for contemplation and actions of decision makers; politicians,
managers, researchers, etc., and the aim was to improve the conditions of these low
resource groups.
During the 70ies, there was a substantial increase in documentation of living
conditions and we find a variety of approaches, from anthropologically inspired
documentary studies of working life (Terkel, 1972) to action research documentaries
(Herzog, 1972), based in, or inspired by, ethnography. Ethnography has also found its
way into the Participatory Design tradition and has enhanced system development
since the 80ies. (Blomberg, Giacomi, Mosher & Swenton-Wall, 1993).
The Scandinavian approach to system development has been dominated by
Action Research methods since the early 70ies. The approach had its basis in the
Scandinavian tradition of collaboration and negotiation on the labour market where
goals were developed in co-operation with Union representatives. During the 90ies,
new participatory methods have evolved focussing on the design of technologies that
would respond to the lived practice of the users. Today, "developing with users" and
"knowing your users" are becoming competitive factors and ingredients in the
business strategy. This is not only true for the new Web-industry, but also for IT-
companies developing more traditional, administrative systems.
During the same period, there has been some major changes in the basis for
design. The focus of the application has changed, and systems are not only developed
to automate low-order skills and routine work. On the contrary, knowledge work is
targeted and the aim is to integrate technology to support and develop complex and
very advanced working processes. The business IT-strategy has changed, and systems
which radically alter the structure and the organisation are being developed and
implemented - a development which may result in radical changes in the value chain.
The technologies have become more complicated and integrated. Stand-alone systems
have changed into integrated, co-operative, distributed and networked systems.
Simple text-based interfaces have become complex, multi-modal symbolic
representations integrating multimedia and 3D, and as systems become more complex
and complicated they require the inventive and the intelligent co-operation of the
users.
Within the system design research community, this development raises
questions about the methods which evolved during the critical era of Action Research
and Participatory Design. Can the same methods be used when the overall aim for the
work as well as the user groups (or domain experts) are changing? What can we bring
with us from the history of Action Research and what requires further development?
Which specific techniques and tools may be used, and equally important, how should
they be organised? And how may we theoretically conceptualise the design methods
we are using?
We try to deal with these questions based on the experiences gained in the
MANICORAL-project. We do this by taking a closer look at the tradition of Action
Research in working life and in everyday life. This will allow us to determine the
differences in:
- the roles of the action researchers
- the methodological approaches
- the roles of the participants
Based on the findings, we will then turn to the methodological framework and
the techniques and tools used in MANICORAL. We suggest Dialogue Design as a
unifying concept and approach, which reflects the needs of a project such as
MANICORAL. The presentation of the methodological framework will allow us to
describe the initial three phases in the project and the techniques and tools we applied
to enhance dialogue. In a final paragraph, we will conclude by reflecting on and
discussing the theoretical conceptualisation of Dialogue Design.
3. FROM ACTION RESEARCH TO DIALOGUE RESEARCH
AND PARTICIPATORY DESIGN
Action Research started in Norway and seriously gained ground in the
Scandinavian research world in the late 60ies and early 70ies (Nygaard & Bergo,
1974). Through close collaboration, the method was adopted in other Scandinavian
countries. Gradually, it became a refined and powerful political tool in the hands of
progressive researchers co-operating with weak resource groups. Despite the fact that
the borderlines are fuzzy, one could say that Action Research took two roads in the
70ies. One was Action Research in working life aimed at professional resource
building (Bansler, 1987), and the other concerned broader aspects of living and
working conditions of the workers. (Danielsen, Drewes & Nielsen,1995)
3.1 Working Conditions
Action research projects with focus on work sites gradually developed the aim
of empowering the workers, through education of the union representatives. The
belief was that technology - if not based on critical reflections on consequences -
would render the professional qualifications obsolete, and threaten jobs. Behind this
approach of professional resource building, was the understanding that the workers
could become qualified players in the technological power game if empowering their
representatives through knowledge and understanding.
During this period of political awareness, co-operation between unions,
researchers, computer scientists and students increased significantly. Unions funded
research and supported a wide range of academic projects. However, many of the
projects within the professional resource building strategy were defensive in relation
to the impetuous of data technology. Gradually, a new strategy of alternative
technology evolved. The aim of this strategy was to develop tools for and together
with the workers. The underlying agenda was technology design which would
enhance the workers' professional qualifications and even enhance living conditions.
In Great Britain, the alternative production as conducted by Mike Cooley, a former
engineer at Lucas, became prototypical. In Denmark, the strong movements in
alternative technology, e.g. the UTOPIA-project, where computer scientists and
typographers worked together, are examples of this. (Cooley, 1987; Ehn 1988)
Prototyping strategies were developed and resulted in a close co-operation
with users. This also ensured an offensive and critically constructive approach, and
alternative technology became a transition concept from action research to
participatory design. (Greenbaum & Kyng, 1991). During the 90ies, Participatory
Design went through several transformations, and one direction seems to be towards
functionalistic action research, where the primary focus is on systems which "must"
work. (Kensing, Simonsen & Bødker, 1998). This change is also due to the
technological development from simple data processing to multimedia, netbased
communication, etc. It challenges R&D within system development and design, where
new methods for user participation, new tools and a new theoretical foundations have
to be developed.
3.2 Living Conditions
The basic approach in Dialogue Research appeared from a focus on, and belief
in, a dialogue between experts and laymen at so called dialogue conferences. The
laymen saw experts as resources. Through questions to and discussions with expert
panels, the laymen would acquire sufficient understanding and knowledge to unfold
their own recommendations to local politicians, administrators, governments, etc.
concerning a given policy. The slogan was: This is much too complicated to hand
over to experts. Future workshops (Jungk & Müllert, 1986) and scenario workshops
were among the techniques used, and the laymen group was composed to represent
the population of a given community: officials, politicians, citizens, school children,
people from supply companies, entrepreneur enterprises and the financial sector, etc.
(Ruus, 1984; Remmen, 1990).
The conception of the process was dialogue understood both as the fundamental
tool, and as the process through which mutual understanding can be reached. The role
of the researcher was to act as midwife for the process, to help with co-ordination, to
set up dialogue workshop and communicative ethic rules so that the participants could
discuss and negotiate between them. S/he also had to act as a critical partner (expert)
in the project but without taking control.
Dialogue Research as a method is also embedded in proactive technology
assessment - not as a traditional product evaluation but as a dynamic process
assessment taking place during the course of a project. This implies that the researcher
is in a continuous dialogue with the acting participants, and it also includes a
continuous presentation and discussion of findings to ensure that the acting
participants are able to influence and guide the process. The locus of control and of
influence is in the hands of the acting participants. In Denmark, the production of
windmills can be seen as a very concrete example of proactive technology assessment.
It has become a profitable business and has developed a stronghold position on the
world market.
A dialogue researcher does not participate in the experimental work but
observes the processes and reports the observations to the actors - as opposed to the
action researcher who takes action together with the actors. One could say that the
role of the researcher is that of a scribe keeping record, and that of a storyteller
recounting the ongoing process. This is the very essence of Dialogue Research. The
methodology was, however, also a deliberate attempt to mark a distance to the critical
action research, which was criticised for being too little research and too much action
(Aagaard Nielsen, 1996). Setting up fora for dialogues and acting as negotiator for the
different interests within a project became a task for the researcher. In this work, the
ethical principles of discourse in Communicative Actions (Habermas, 1991), became
the epistemological inspiration for the approach. (Duelund, 1991).
Figure 2: General outline of the development of Action Research from the 1970ies to
the 1990ies.
"FIGURE 2 INSERT HERE"
Figure 2 illustrates the two directions of Action Research in the early 70ties and
how the two strands: Participatory Design and Dialogue Research have developed and
have been brought together again in Dialogue Design in the late 1990´ies. (Danielsen,
1997).
In MANICORAL, the process of developing and testing tools and techniques
which would enhance dialogue and mutual learning alternated in a dialectic process
with the development of the methodological framework of Dialogue Design. In this
paper, we have chosen to communicate this very complex process by introducing the
methodological framework of Dialogue Design which also embeds the principle of
mutual learning. This presentation will allow us to move directly to the description of
the three phases and to the techniques and tools applied during the course of
MANICORAL.
4. DIALOGUE DESIGN
As MANICORAL evolved, it became clear to us that in order to open for and
enhance co-operative processes, we had to put special effort into the communicative
acts and into the processes of mutual learning among participants. It had to be
reflected that the project represented a special kind of action research due to the fact
that:
- It was a truly international, inter-cultural and interdisciplinary project between
equal groups
- All participants were knowledge workers and high resource groups with regards
to educational and technological skills
- The design should support the practice of researchers from the Natural Sciences
collaborating on complicated research - and themselves being expert
programmers
Our answer to this diversity became methodological – drawing on Participatory
Design (PD) and Dialogue Research (DR). In DR, it was the fundamental
understanding of dialogue as a dynamic process unfolding during the course of a
project which became essential for the communicative actions and for experimenting
with different prototypes in MANICORAL. In PD, the aim of understanding practice
is fundamentally tied to informing and creating design, and to do this together with
users. This understanding became essential with the aim of developing a DCV-system
together with expert users. The approaches were further enhanced through our focus
on mutual learning as a necessity in a multi-cultural group. The unifying concept -
which also became the methodological frame - was Dialogue Design (DD) and the
techniques and tools evolved from the day-to-day work in MANICORAL.
However, the methodological framework of DD differs from the above in a
number of essential ways. Where Action Research historically favoured the
underprivileged groups, DD has to support knowledge workers. As domain
specialists, they are high resource groups (regarding educational level, technical skills,
job positions, etc.), and are themselves developers of new technology – they have a
history of writing their own software. They are not “typical” users. On the contrary,
there was a symmetric relation between all the actors involved in this design process:
The users, the HCI-researchers and the technical experts. They all had academic
backgrounds and were involved in research, and they were all experienced users of
technology. Besides, they had defined the project, joined it by own wish and received
economic support to cover part of their participation.
The implication of the above statement is not that only academic players may
participate in Dialogue Design projects. Our understanding is that a methodological
framework may serve many different projects but the techniques and tools must be
iterated through careful understanding of the client, the users, the organisation, the
task, etc. And sometimes the understanding acquired may lead to refutation of the
framework. There is no "one-size fits all". Projects differ significantly, users differ
significantly, clients differ significantly, the conditions under which the system is to
be developed differ significantly, the economic situation differs, etc. Each new system
project requires careful reflection and investigation before decisions are made and the
decisions cannot be final. The process is iterative, participatory, and the aim is a
functional and pragmatic design which responds in a visionary way to the daily work
practices of knowledge workers.
4.1 Dialogue - A Mutual Learning Process
The concept of dialogue is, like the concept of Dialogue Research borrowed
from Jürgen Habermas and his theory of communicative action, (Habermas, 1991).
The communicative action is understood as a true dialogue between “rational arguing”
participants in terms of comprehension, truth, rightness and thrustworthiness. Ideally,
there are no hidden agendas, and the participants meet with open minds and with the
understanding that the best argument will win.
“The concept of communicative action refers to the interaction of at least
two subjects capable of speech and action who establish interpersonal relations
(whether by verbal or by non-verbal means). The actors seek to reach an
understanding about the action situation and their plans of action in order to
co-ordinate their actions by way of agreement (ibid. p.86).
…. An actor who is oriented to understanding in this sense must raise
at least three validity claims with his utterance, namely:
- That the statement is true .. and that the content mentioned is in fact satisfied
- That the speech act is right with respect to the existing normative context .., and
- That the manifest intention of the speaker is meant as it is expressed" (ibid.
p.99, our accentuation).
The ideal communicative action requires that the dialogue process is a learning
process. In a "true" dialogue, participants will challenge each other’s fundamental
understandings of truth, norms and trustworthiness, because of conflicting
perspectives and experiences. As such, dialogues are the means to rearrange, renew,
and reorganise fundamental assumptions. Learning is therefore fundamentally
embedded within communicative actions.
“The concept of argument is closely connected with the concept of
learning …. Rationality is only random as long as it is not coupled with the
ability of learning from mistakes, learning from the confutation of hypotheses
and the failure of interventions”. (Habermas, without year, p. 44, own
translation).
This implies that when participants challenge each other’s basic understandings
through various dialogue techniques and reorganise the fundamental assumptions and
perspectives, then mutual learning takes place. Mutual learning is therefore the other
key concept in Dialogue Design. Argyris (Argyris, 1977) makes a distinction between
single loop and double loop learning, which seems to correspond to the distinction
Piaget (Piaget, 1992) makes when he speaks of assimilative and accommodative
cognitive processes. Single loop learning and the assimilative cognitive process means
that the learner adjusts and corrects, yet does not challenge her/his fundamental
understandings. Instead, new information is transformed and fitted into already
established schemes and understandings. Double loop learning and accommodative
cognitive processes, however, truly challenge the established understandings, and
change the learners' basic assumptions. Thus existing schemes and understandings are
accommodated to fit the new insights. Moreover, double loop learning and
accommodative processes embed the cognitive qualifications, which make critical
reflections (and self-reflection) possible. Here, we enter into triple loop learning
which may be explained as thinking about thinking. Where single loop learning is
reflection in action, double loop learning is reflection on action and triple loop
learning is reflection on reflection on action.
Figure 3: The triple loop learning model
"FIGURE 3 INSERT HERE"
Dialogue Design, as a method, aims at double loop and triple loop learning
processes. The ability to reflect and accommodate own understanding and goals in the
project to those of the other groups has already been pointed out. But also the ability
to reflect on own scientific understandings and Weltanschauung - triple loop learning
- is a must in multi-cultural projects like MANICORAL. The participants came from
eight different countries, they each had their own language, cultural heritage and
customs. At the same time, they carried with them the very different understandings
and world views embedded in their scientific disciplines. Accommodation to new
insights as well as meta-reflection was necessary if the project was to succeed. But
triple loop learning processes were also required because the design had to be oriented
towards an unknown future - towards that which is not. It had to be visionary and the
challenge was to re-assess the conditions for the design, to imagine what did not yet
exist and to adapt to new ideas and visions.
5. PRACTICES
With the presentation of the methodological framework on which the work of
the HCI-group became based, we now move into describing how opportunities for
dialogue, interaction and mutual learning processes were designed in the project. We
do this by providing a general introduction to three phases in the project and by
presenting different techniques applied during the different phases. Methodologically,
the projects may be divided into three practice phases:
- Ethnographic and qualitative methods to understand practice
- Constructive design methods in order to produce a vision for practice
- Negotiation of design options in order to reach decisions about practice
5.1 Understanding Practice –Baseline Data
The design of the DCV-system had to be closely integrated with everyday
research practice of the domain experts, and had to build on their experiences.
Requirements must derive from a deep understanding of practice and this
understanding cannot be acquired primarily through detached contemplation. To
acquire understanding of life, it must be lived and experienced before its depths can
be contemplated. This is the essence of the ethnographic approach (Høyrup, 1983),
which has also found its way into design, where a general framework served as
inspiration for our work (Blomberg et al., 1993; Trigg, Anderson & Dykstra-Erickson,
1993). This was elaborated with a critical hermeneutic approach (Nielsen, Dirckinck-
Holmfeld & Vendelø, 1996) because doing research, not merely through detached
contemplation but also by living life, means that we, as researchers, are emerged in
interpretative work. Designing investigations, applying theory, collecting empirical
data, and expressing them in symbolic representations are all processes of
interpretation. However, interpretations need to be made objects of detached
contemplation and critically reflected upon, and methods and techniques for
validation must be applied.
The field studies were conducted using different techniques such as field
observations including video recording, still life photography transformed into
photojoiners, informal interaction techniques, semi-structured interviewing techniques
and document collection and analysis. The focus was on the following activities and
practices:
- Mutual interviewing among all participants on the understandings and goals
related to the project
- Field studies of daily practice in the different local research communities within
the AFRICAR-community
- Studies of physical meeting cultures in the AFRICAR-community
- Studies of virtual meeting cultures in the AFRICAR-group mediated by CSCW-
tools
- Mind tapes to reflect on the experiences from and the observations of the virtual
meetings
- Mutual teaching sessions
To enhance mutual learning and dialogue, the study results were reported to the
participants in the project using various dissemination techniques: presentation and
discussion of data at project meetings, discussion sessions where video sequences
from virtual meetings were replayed for the user group, and finally, mind-tape
sessions with users (Nielsen & Christiansen, 2000). The function of these were to
enhance understandings, to validate or refute interpretations, and to help identify
communication problems and co-operation needs.
These baseline studies contributed to an understanding of:
- the other participants' motives and goals related to the project
- the different cultures among the main groups in the project
- the potential conflicts and interests in the project
The studies also fertilised the ground for the growth of:
- trust and respect for each other
- establishment of a common ground between the participating groups regarding
ways of working and main methodological approaches
- evolving consciousness and acceptance of the scientific and cultural differences
in the groups
Finally, these techniques also contributed to:
- a grasping of the experiences of the domain experts
- a transferral of requirement captures from the domain specialists and the HCI-
researchers to the system designers
- the evolving of a mutual learning culture based on an interest in learning from
each other
A central task was transferral of requirements. We found that video techniques
may be used constructively and push transfer between the users, the HCI-specialists
and the system developers and programmers. (Dirckinck-Holmfeld, 1997).
Traditional requirements capture is focused on validating and documenting the
existing practice in written documents. However, video recordings may be used as
documentation and as material for communication, and as “teaching material”
regarding critical design issues and may also be used to create stories and scenarios
for design. For the time being, these possibilities have not been further explored and
documented but it is quite clear that visualisation enhances understanding and easily
promotes discussion of critical issues.
5.2 Visionary Practice – Forays of Dialogues
Developing a CSCW-tool and a DCV-tool for a group of domain experts means
developing for an unknown future. Rich visions about that future are important, as
there are no straightforward solutions. Therefore, a number of scenarios was
constructed. This scenario process continued over time allowing for more focused
visions of the domain experts. The process of working with scenarios is very essential
as borders of the possible/impossible in technology are pushed and the visions also
gain by gradually becoming clearer.
Requirements capture without opening for visions among the domain experts
may result in requirements that are too conservative. As such, we see the HCI-
researcher as a sort of change-agent, and the approach as basically a change-
methodology - changing the present for a possible future.
In order to be able to produce rich visions and to share the visions in the group,
we used different techniques:
- Mind mapping (used in the initial phase of conceptualising the project)
- Mutual interviews in order to grasp the dreams and visions of all the participants
- Future workshops to formulate the first visions for the project
- Training sessions to learn about the possibilities within the new technologies
and to learn about human communication and collaborative strategies
- Experimental sessions: implementing and trying out the CSCW prototype with
the users in their natural research setting
- Scenario design sessions
These visionary practices contributed to an understanding of:
- the many different visions for the project
- the different expectations to the future collaboration among the local user
groups
- the difference between virtual and physical face-to-face interaction
The studies further enhanced the growth of:
- "Wild" ideas
- Trust in and respect for each other
- The common ground between the participants and the diversities
- The accept of different priorities and different interests
- The willingness to compromise and to fight
Finally, the techniques also contributed to:
- the opening of discussions among the domain experts regarding different kinds
of co-operation (conveyer-belt as opposed to dialogue)
- narrowing in the ambitions for the entire project
- closer collaboration between users and technicians
- the continued evolvement of the mutual learning culture
5.3 Decision Techniques – From Requirements to Specification
In the MANICORAL project, we wanted the requirement specification to be a
result of a mutual dialogue between the different actors: The HCI-group, the users and
the technical experts. In other words, the specification really had to be a melt between
the different perspectives and possibilities but with the users' needs and visions as a
driving force. This was a very difficult process. We were dealing with very
sophisticated choices to be made between different groups which were experts in each
their field: in communication and co-operation, in the geophysical domain, and in
CSCW-supporting technologies. Each group had acquired profound knowledge
within their field, and were able to point to problems and suggest solutions. But they
needed the knowledge of the other groups in order to come up with really viable
solutions. Hence none of the groups were in a position where they could make
decisions independently, or on behalf of the others. The choices to be made had to be
founded on “the better argument”. In order to assist that process, we were inspired by
different approaches to decision making (Enderud, 1976), and the basis became:
- Reports and analyses of findings
- Participatory design sessions building on scenario design
- Dialogue and decision meetings
The overall aim was to implement some techniques and tools in the design and
in the development process, which supported all the participants in the central
decisions regarding the project. Different techniques may be used to throw light on
which decisions to be made but they also offer alternative ways for making decisions.
Our participatory approach ensured that the design reflected the experiences, the
visions and the needs of the domain specialists and that the solutions were worked
out together across disciplines with due understanding of the communicative needs in
a virtual co-operation and of the limitations that technology imposes.
The most important tool was the participatory design sessions building on
scenario design (Carroll, 1995). The point of departure was general scenario
descriptions of typical collaborative sessions. These descriptions were developed
from the empirical analysis of physical meetings, from interviews and field work,
from the imaginative visions produced, and also from the analysis of virtual
meetings. A typical scenario was drafted and then iterated a number of times in close
collaboration with the domain experts. Once there was agreement to the general
scenario, it was analysed for more specific design implications.
One focus point in the decision process was: Communication and Co-operation.
Should the design be such that speaker control was in the hands of the host of a
meeting, or were other solutions possible, e.g. developing a new meeting culture
among the AFRICAR-researchers, with new ways of interacting? Should the design
allow for all participants to be able to change a shared image or should it only be the
owner of the document? Scientific aspects were also essential issues in the design
decisions, e.g.: What was the consequences of not being able to share hand written
mathematical notations. And finally, also technical aspects were in focus, such as:
Was it possible to develop the system in such a way that the DCV-tool was
seamlessly integrated with the different databases to which the geophysicists needed
access?
These data together with the general scenario were then divided into design
issues, which were presented at a design workshop where the participants discussed
possible decisions in interdisciplinary groups. The decisions of the different groups
were presented and discussed at plenary sessions, and general agreements concerning
the design were then taken.
The data material was handed over to the group in charge of the development.
This group consisted mainly of system developers and programmers but two
geophysicists from AFRICAR and two HCI-people also participated and met on
regular basis to discuss progress and problems (Nielsen & Lindgaard, 1997). This
approach allowed for very visionary wishes to be juxtaposed with actual practice and
related to virtual experiences and equally important to the technological possibilities
and constraints.
The above mentioned types of methods and techniques (Nielsen, Dirckinck-
Holmfeld & Vendelø, 1996) made up the basic structure within Dialogue Design. To
the HCI-researchers, the richness of techniques and tools as well as the way in which
the methods were structured were very productive for a visionary design. The
richness ensured that a phenomenon was worked upon from many angles but built on
the experience of the domain experts. However, the different communication
techniques also challenged the domain experts on how to communicate and
collaborate and this stimulated an expansion of their ideas and visions for the
distributed collaborative visualisation system (DCV). Sometimes far more than the
system developers could technically comply with.
Before we move on to the discussion and contemplation of our methodological
framework, we want to give an example of how dialogue and mutual learning
constituted the basis. The following account is drawn from the initial phase of
understanding practice, and we describe the creation of a forum for mutual learning
processes based in the communicative act.
6. MUTUAL INTERVIEWING AS A TOOL FOR DIALOGUE
AND LEARNING
The beginning of the MANICORAL-project was an invitation to a meeting to
four research communities (Natural Sciences, Social Sciences, Humanities and
Technology). They all sent members to participate in a two-day workshop, the aim of
which was to explore the possibilities for a joint CSCW-project, and the possibilities
for establishing funding within the EU (European Union, R&D-programs). Through
presentations, and by using Mind Mapping as a tool for generating ideas and
extensive discussions, the project was conceived, and a rough framework defined. A
committee with representatives from each of the four sciences participating was then
established and was responsible for writing the proposal. They were also, each of
them, responsible for updating and discussing the progress of the proposal within each
their research community.
Once the proposal had been sent to the EU, there was a long waiting period and
many of the reasons for the finer arguments or specific paragraphs put forward in the
proposal were forgotten. Also the technological development and the research within
the different scientific groups had developed in different ways, and many things had
changed. When the invitation to go to Brussels to meet with the EU-officers finally
arrived, more than a year had passed since the initial workshop. At the meeting, the
EU reduced the project costs with 50%, and the project proposal had to be rewritten
and changed into a contract – in accordance with EU-regulations for R&D-proposals
under the 4th frame program.
The contract was eventually signed, the administrative procedures implemented
at all the participating institutions, and the project could finally begin. But, by now,
more than 18 months had passed since the workshop where the project was conceived
and the engaging project description had been reduced to a contract of formalities –
unrecognisable and incomprehensible to many. Even the authors of the proposal were
in need of re-interpretations in order to understand the text. Besides, the local research
communities originally participating in the initial conceptual phase, were “the old
boys and girls” who all knew each other. When the project started, however,
newcomers, PhD-students and assistant professors, had joined, neither of who had
participated in the conceptualisation phase.
The HCI-researchers' solution to this diversity was to start off the first meeting
by asking the participants to interview each other. First, the members from each of the
four research fields worked together on their understanding of the project, and on
specifying what they were going to do in the project. Then the disciplinary groups
were split up and interdisciplinary groups were formed. Each researcher now
represented his/her research field and the task s/he had was to interview the other
representatives regarding their understandings of and their goals related to the project.
Finally, the interdisciplinary groups had to reach an agreement and present their
definition of the project and common goals and goals of local groups at a plenary
session.
Though the HCI-researchers expected some degree of diversity, the extent of the
diversity came as a surprise to everybody in the project. However, the exercise proved
essential for getting the participants to participate in a dialogue about the
specifications in the contract, their perception of the project goals, related to those of
their individual research groups and to the goals of others. The discussions and reports
immediately made it clear that a balance had to be reached and identification of
common tasks and goals became a must. But more than anything, the interdisciplinary
mutual interviewing and the process of prioritising ensured that all participants
became aware of the importance of listening to the others, of asking questions and of
following up on answers in order to ensure constructive thinking in MANICORAL as
a whole, as well as within individual groups. The important goal of the exercise was
to demonstrate the diversity and to assist the first steps towards developing common
ground and to allow for the development of ownership.
This technique applied at the very first meeting also set the agenda for the
following meetings. Though we did not initially realise the cultural impact of this
technique on the project, it quickly became an accepted approach that extensive
dialogue and the need to listen carefully to the others was the way of reaching
agreements in MANICORAL. This is not to say that the process was easy and
smooth, on the contrary. Listening is not the same as understanding, and the second
approach in the project, that ofmutual teaching almost suggested itself.Creating forum
for the different groups to "teach" each other, to present their research, to explain to
others the what, why, how of their field, and to discuss with researchers from other
disciplines established a space for learning in the dialogue. .
7. DIALOGUE DESIGN AS MUTUAL LEARNING – SUMMING
UP ON THE PRINCIPLES
On the basis of our experiences gained in the MANICORAL-project, we can
summarise the principles for Dialogue Design in the following statements:
- The fundamental tool in DD is the dialogue and a basic competence within
mutual learning
- DD is fundamentally a mutual learning process between equal professional
groups who have distinctly different tasks to perform and different roles in the
design project
- DD, at present, focuses on the following three phases:1) understanding practice,
2) create visions for practice 3) negotiate decisions about practice
- Communication, learning processes and decisions take place within a field of
many different life worlds and practices. This complexity of cultural, scientific
and methodological differences should be thematically approached during the
design process
- Dialogue is not only about finding the “common denominator”. The process
must be critical of the system and self-critical in relation to own practice
- The communicative process demands that fora for dialogue and mutual learning
processes are constructed, and multiple communicative codes are ensured
enough space
Participating in an R&D-project, where dialogue and mutual learning processes
are the fundamental principle, challenges the HCI-researcher. In the design process,
the role of the HCI-researcher is, generally speaking, to act as midwife for the
complicated process of communicative action and mutual learning. The HCI-
researcher needs:
- to establish different types of fora for dialogues and mutual learning using
different modes of communication and interaction
- to have a deep understanding for different discursive communities
- to seek a mutual understanding of the design by trying to capture the “ways of
reasoning” behind the competing views – without devaluing or excluding a
priori
- to acquire insight into the task domain
- to establish communicative legitimacy to reconstruct and to present the interests
of the different partners
- to be able to mediate the negotiation process between conflicting interests
8. REFLECTIONS ON THEORETICAL FRAME
The principle of Dialogue Design is an ideal. It is based on the normative
premises that it is possible, through the ideal of communicative action, to critically
process and exceed the power relations and cultural boundaries that exist in
interdisciplinary, intercultural, cross-national, cross-generational and cross-gender
groups as in MANICORAL. Or to voice it differently: Because of the heterogeneity of
the project it is essential to focus on the dialogue and the conditions for mutual
learning in order to bring out new design. But it is very hard work to reach design
agreements.
The basic premises for Habermas' dialogue and learning concepts is rationality,
as the organising principle in the communicative action. To a large degree, we accept
this ideal as it points, in its utmost consequence, towards a vision full of hope. The
better argument takes precedence and a future of peace, a society where dialogue, not
war and weapons, is the tool for solving conflicts truly shows a (very high) level of
cultural development. However, every day reality – also Dialogue Design processes–
show that this is a very idealistic approach. People are not always rational, they have
different agendas, they are also emotional and they react in affective ways. Besides,
the conditions for communication without domination are not often present. In the
case of MANICORAL it was, however, necessary to establish an ideal about
communicative action. Partly because research is about seeking the truth, and partly
because of the very interdisciplinary nature of the project: How can we decide
between the different interests in such a project without leaning on the better
argument?
In Habermas' communicative action, the basis is the rational discourse in which
explicit verbalisation is embedded. However, in DD, the visionary work is important:
to imagine the impossible. In order to come up with something really new, one has to
move beyond the structuring rationality and the rational use of language. One has to
open for other ways of seeing and discovering - one has to allow for the “wild”
imagination and for the tacit knowledge to take over, and not be constrained by
symbolic representations such as language (Nielsen, 1996). As a consequence, the
communicative dialogue is in itself insufficient. Other ways of knowing (Polanyi,
1968) and a multitude of communicative forms must be included when dialogue
researchers wish to construct dialogue workshops that may open for new possibilities
and “move” the design process. In the methodological frame in DD, we have taken
this into consideration in several ways. We have worked with vision techniques, used
different aesthetic and kinetic communications means, used photography as a way of
documenting cultures. We have "mis-en-scene" non-verbal interactions among
participants as a tool for creating awareness and used cartoon drawings as a way of
explaining things, which could not otherwise be made explicit. However, it is beyond
the scope of this paper to document all the techniques, this is for future work.
NOTES
Acknowledgement: In the beginning of the project, the technological experts
from Rutherford Appleton Laboratories sweetly pointed out to us that “these projects
always end in tears”. Yet, we never cried. By the end of the project, the geophysicists
pointed out – with a warm and apologetic little smile - that the project had taught
them that “what social scientists and humanists do is also science and research”. In
return, we thank you all for opening the wonderful world of your scientific fields for
us, for the good dialogues, the good fights, the patience and the many laughs – we
could not have wished for better partners.
Drafts of this paper have been discussed in the HCCC-group in MANICORAL and in
the Danish research network Learning and Multimedia. We thank our colleagues for
critical and constructive comments. As always, a special and final thank you goes to
research secretary, MA, Hanne Porsborg Clausen. Again and again Hanne steps in an
saves us with her expertise, her sense of responsibility and her profound
professionalism.
REFERENCES
Argyris, C. (1977) Double-loop learning in Organizations. Harvard Business Review,
55 (5) 115-125
Bansler, J. (1987) Systemudvikling – teori og historie in Skandinavisk perspektiv.
Studentlitteraturen, Lund – Sweden
Blomberg J., Giacomi J., Mosher J. & Swenton-Wall, P. (1993) Ethnographic Field
Methods and their Relation to Design. In Schuler, D. and Namioka, A. (eds.):
Participatory Design: Perspectives on Systems Design. Hillsdale, New Jersey:
Lawrence Erlbaum Associates: 123-154.
Bødker, K., Simonsen, J.; Kensing, F. (1997) MUST - en metode til forundersøgelse
med brugerdeltagelse in Fibiger, B.: Design af Multimedier, Aalborg University Press,
19-60
Caroll, J. (1995) Scenario-Based Design. Envisioning Work and Technology and
System Development. John Wiley & Sons. Inc. New York
Cooley, M. (1987) Architect or bee? The Human Price of Technology. Rev. ed.
Hogarth Press, London
Danielsen,O. & Drewes Nielsen,L. (1995) Scenario Workshops on Urban Ecology in
Four Local Communities, The European City – Sustaining Urban Quality, ,
Conference Papers vol. 3, European Commission DGXI and Ministry of Environment
and Energy April 24th-28th , Copenhagen, 1-7
Danielsen, Oluf (1997) From Action Research Via Dialogue Research to Dialogue
Design, in Dirckinck-Holmfeld, L. (ed.): Dialogue Design - Methodological
Considerations in the MANICORAL-project. EU-RE 1006, deliv. O5.8.2, 7-14
Dirckinck-Holmfeld, L.(1997) The Video as a Tool in Dialogue Design, in Dirckinck-
Holmfeld (Ed.): Dialogue Design – Methodological Considerations in the
MANICORAL-Project. EU-RE 1006, deliv. 05.8.2 , 15-23
Duelund, P. (1991) Dialog Research Theories and Methods, in Cronberg, T., Duelund,
P., Jensen, O.M. & Qvortrup, L. (Eds): Danish Danish Experiments - Social
Constructions of Technology, New Social Science Monographs, Institute of
Organisation and Industrial Sociology, Copenhagen Business
School, Copenhagen, 167 - 185
Ehn, Pelle(1988) Work-Oriented Design of Computer Artifacts. Arbetslivscentrum,
Stockholm
Enderud, Harald(1976) Beslutninger i organisationer, Fremad, København., 10th
edition
Greenbaum, J. & Kyng, M. (1991) Design at Work, Cooperative Design of Computer
Systems, Lawrence Erlbaum Associates, Publishers
Habermas, J. (1981)Teorien om den kommunikative handlen (uddrag), Aalborg
University Press 1996. Translated from Jürgen Habermas: “Theorie des
kommunikative Handlens”, Bd. 1-2, Suhrkamp Verlag, Ffm.
Habermas, J (1991) The Theory of Communicative Action, Vol. 1, Polity Press (First
published as Theorie des Kommunikativen Handels, Band 1: Handlungsrationalitat
und gesellschaftliche Rationaliserung, 1981 by Surhkamp Verlag, Frankfurt am Main.
Translation copyright and first published by Beacon Press 1984)
Herzog, Marianne(1976) Von der Hand in den Mund, Frauen im Akkord, Rotbuch
Verlag, Berlin
Høyrup, T. (1983) Det glemte folk, Livsformer og centraldirigering, Statens
Byggeforskningsinstitut, Hørsholm
Jungk, R. (1986) Håndbog i Fremtidsværksteder. Politisk revy. 2. ed. Copenhagen
Lewin, K. (1975) Field Theory in Social Sciences - selected theoretical papers. (Ed)
Dorwin Cartwright, Reprint. Wesport, Conn. Greenwood (original 1952)
Nielsen Janni and Nina Christiansen (2000) Mindtape – a tool for reflection in
Participatory Design in Proceedings for PD, New York, pp. 309-314
Nielsen, J., Dirckinck-Holmfeld, L. & Vendelø, M.(1996) Conceptual Framework -
Methodological Framework for MANICORAL, Copenhagen Business School
Nielsen J., Duce D., Knudsen P., Sünkel, H. and Robinson, K. (1995) MANICORAL
Multimedia and Network In Cooperative Research And Learning, project contract,
EU, 4the Telematics program. RE 1066
Nielsen, J. & Lindgaard, G. (1997) The Use of Scenarious in Dialogue Design, in
Dirckinck-Holmfeld (Ed.): Dialogue Design – Methodological Considerations in the
MANICORAL-Project. EU-RE 1006, deliv. 05.8.2
Nielsen Janni, Lindgaard Gitte, Dirckinck-Holmfeld Lone, Vendelø Morten,
Danielsen Oluf and Marianne Georgsen(1997) User requirements capture for a
multimedia CSCW system, Proceedings of INTERACT, Sydney
Nielsen, K.A. (1996) Arbejdets sociale orientering, Forlaget Sociologi
Nygaard, K&Bergo, O.T. (1974) Planlegging, styring og databehandling: en
grunnbok for fagbevegelsen, Tiden Norsk Forlag,
Piaget, J.( 1992) Barnets psykiske udikling. Hans Reitzels forlag, Copenhagen..
Translated from French: “Six Etudes de Psychologie”, Paris 1964. (In Danish,
1969/1971/ 92)
Ploger, J. (1998) Kommunikativ Planlegging, NIBR-notat 1998:106, Norsk
Institut for by- og regionsforskning, NIBR, Blindern, Oslo
Polanyi, Michael (1967) Logic and Psychology, American Psychologist 23, 27-4
Remmen, A. (1990) Constructive Technology Assessment in Cronberg, T., Duelund,
P., Jensen, O.M. & Qvortrup, L. (Eds): Danish Experiments – Social Construction of
Technology. Institute of Organisation and Industrial Sociology, Copenhagen Business
School, Copenhagen, 185 – 200
Ruus, H. (1984) Danskere for fremtiden, Dansk Dataforening, Copenhagen
Trigg, R., Anderson, S.I. & Dykstra-Erickson, E. (Eds.)(1994)PDC'94: Proceedings of
the Participatory Design Conference. Palo Alto, CA: Computer. Professionals for
Social Responsibility
Tscherning C.C., Nielsen Janni, Arabelos D., Haagmans R., Knudsen P, Sanso F. and
H. Süenkel (1997) Computer Supported Co-operation – the future of the IAG
Information Service and of co-operation within IAG. Proceedings of International
Associations of Geodesy, Scientific Assembly, Rio de Janeiro, September
Terkel, Studs (972) WORKING People Talk About What They Do All Day and How
They Feel About What They Do, Avon Books, New York
figure no. 1
HCCC - commmunication - psychology - social sciences - (Denmark)
Geophysicists - theoretical mathematics - ocean tides& circulation - satellite position - satellite orbits - gravity (Italy, Greece, Denmark, Austria & Netherlands)
- visualisation - internet - cscw - (Great Britain)
Humanities and Social Sciences
Technology Natural Sciences
Figure no. 2
1970s
Action research
Professional resource building (research in working life)
Action research on
Alternative technology Dialogue research (participatory research/evaluation research/proactive technology evaluation)
1980s
1990s
Participatory design
Dialogue research on living conditions (policy research)
Dialogue design
Figure no. 3)
Triple loop
Double loop
Single loop
Reflection
Experience
Theory
Action
Experience
Reflection
Theory
Action
Experience
ReflectionTheory
Action